The paper "Beyond Discrimination: Generative AI Applications and Ethical Challenges in Forensic Psychiatry" by Leda Tortora explores the transformative impact of generative artificial intelligence (GenAI) on forensic psychiatry and criminal justice. GenAI, which can generate realistic data and integrate unstructured content, marks a significant shift from traditional discriminative AI models. The paper reviews the current applications of discriminative AI in forensic psychiatry, such as risk assessment, diagnostic support, and treatment planning, and discusses the potential of GenAI to transform these practices through multimodal generative models, data generation, and data augmentation. Key topics include the development and types of large generative AI models (LGAIMs), such as Generative Adversarial Networks (GANs), Transformer-based models, diffusion models, variational autoencoders, and neural radiance fields. The paper also examines the ethical and legal challenges associated with deploying GenAI, including biases, transparency, data privacy, intellectual property rights, and overreliance on AI outputs. It highlights the need for interdisciplinary collaboration and thorough evaluations to address these challenges and ensure responsible and ethical use of GenAI in forensic contexts. The paper concludes by emphasizing the dynamic challenges and opportunities presented by GenAI in forensic psychiatry, advocating for careful consideration and regulation to prevent potential misuse and ensure fair and just decision-making processes.The paper "Beyond Discrimination: Generative AI Applications and Ethical Challenges in Forensic Psychiatry" by Leda Tortora explores the transformative impact of generative artificial intelligence (GenAI) on forensic psychiatry and criminal justice. GenAI, which can generate realistic data and integrate unstructured content, marks a significant shift from traditional discriminative AI models. The paper reviews the current applications of discriminative AI in forensic psychiatry, such as risk assessment, diagnostic support, and treatment planning, and discusses the potential of GenAI to transform these practices through multimodal generative models, data generation, and data augmentation. Key topics include the development and types of large generative AI models (LGAIMs), such as Generative Adversarial Networks (GANs), Transformer-based models, diffusion models, variational autoencoders, and neural radiance fields. The paper also examines the ethical and legal challenges associated with deploying GenAI, including biases, transparency, data privacy, intellectual property rights, and overreliance on AI outputs. It highlights the need for interdisciplinary collaboration and thorough evaluations to address these challenges and ensure responsible and ethical use of GenAI in forensic contexts. The paper concludes by emphasizing the dynamic challenges and opportunities presented by GenAI in forensic psychiatry, advocating for careful consideration and regulation to prevent potential misuse and ensure fair and just decision-making processes.